eLetters

13 e-Letters

published between 2021 and 2024

  • Evaluation of Local Open-Source Large Language Models for Clinical Data Extraction

    Dear Editor,
    I am writing to commend the authors for their insightful manuscript titled "Evaluating local open-source large language models for data extraction from unstructured reports on mechanical thrombectomy in patients with ischemic stroke." (1). This study provides a rigorous evaluation of local open-source large language models (LLMs) in extracting clinical data from procedural reports, an area of increasing relevance given the rise in AI applications in healthcare.
    The manuscript presents a well-structured methodology for assessing the performance of three LLMs—Mixtral, Qwen, and BioMistral—on data extraction tasks from thrombectomy reports. The choice to focus on local models, as opposed to commercial counterparts, is particularly noteworthy due to the enhanced data privacy and security benefits it entails.
    Strengths of the Study:
    1. Comprehensive Approach: The use of a robust human-in-the-loop (HITL) annotation strategy to establish ground truth is a commendable approach. By incorporating expert validation into the workflow, the authors not only enhance the reliability of their results but also address one of the key challenges in data extraction—ensuring the accuracy of the extracted data.
    2. Clear Evaluation Metrics: The manuscript's use of precision, recall, and F1 score as metrics for evaluating model performance is appropriate and provides a detailed picture of each model's efficacy. The precision metrics, p...

    Show More
  • Correspondence on ‘Outcomes of mechanical thrombectomy in stroke patients with extreme large infarction core’

    We read with great interest the corresponding letter by Dr. Sacks regarding our research article about the outcomes of mechanical thrombectomy in patients presenting with ASPECTS 0-2. We appreciate Dr. Sacks’ note regarding the acceptable clinical outcome (90-day mRS 0-3) in our study which was comparable to the outcomes reported in the large core clinical trials recently published. Regarding the mortality rate, it was reported as 4.5% in error. The correct mortality rate for patients with ASPECTS 0-2 who underwent mechanical thrombectomy is 24/58 (41.4%). We have submitted a correction to our article to reflect the correct mortality rate and updated Table 3 and Figure 2 to clarify Questions 3-5.

  • The Importance of Thrombus Composition in Endovascular Treatment Outcomes

    Dear Editor,
    I am writing to express my thoughts on the article titled "Intracranial thrombus composition is associated with occlusion location and endovascular treatment outcomes: results from ITACAT multicenter study", recently published in the Journal of NeuroInterventional Surgery (1). This study provides pivotal insights into how thrombus composition affects the efficacy of mechanical thrombectomy (MT) in treating ischemic strokes, highlighting critical aspects that could influence future therapeutic strategies.
    The study's retrospective analysis of 221 patients, examining thrombi retrieved during MT using hematoxylin–eosin staining and CD61 immunostaining, revealed significant associations between thrombus composition and treatment outcomes. The finding that medium distal vessel occlusions (DMVO) had higher platelet proportions than proximal large vessel occlusions (LVO) is particularly noteworthy. Specifically, middle cerebral artery (MCA) M2–M3 segment thrombi had the highest platelet content (PLTPT at 70%), whereas basilar artery thrombi had the lowest (PLTPT at 41%). These differences suggest that occlusion location can be an indicator of thrombus composition, which has crucial implications for tailoring treatment approaches.
    The logistic regression analysis further strengthened the evidence by showing that lower baseline National Institutes of Health Stroke Scale (NIHSS) scores and lower PLTPT were independently associated with s...

    Show More
  • The Ongoing Debate Over Dual Antiplatelet Therapy in Cerebral Aneurysm Treatment

    Dear Editor,
    I am writing to address the significant findings presented in the article "Short- versus long-term Dual AntiPlatelet Therapy for Stent-Assisted treatment of CErebral aneurysm (DAPTS ACE)." This research, a randomized clinical trial aimed at unraveling the ideal duration of Dual AntiPlatelet Therapy (DAPT) after stent-assisted coil embolization (SACE) for cerebral aneurysms, raises crucial questions regarding the management of these delicate cases.
    The article reports that the trial, DAPTS ACE, investigated whether extending DAPT for 12 months could effectively reduce the risk of ischemic stroke in comparison to a shorter 3-month DAPT regimen for patients who underwent SACE for cerebral aneurysms. The study, conducted in 17 hospitals in Japan, enrolled 142 patients between November 4, 2016, and January 7, 2019, and assigned them randomly to either the long-term or short-term DAPT group. The primary outcome was the occurrence of ischemic stroke events during the period of 3 to 12 months after SACE.
    The results are undeniably important, as they indicate that there was no statistically significant difference in the rate of ischemic strokes between the long-term and short-term DAPT groups. In fact, ischemic stroke was a rare occurrence in both groups, and the data show that the extended duration of DAPT did not offer a significant advantage in preventing ischemic events. This is the first randomized controlled trial to tackle the duratio...

    Show More
  • Comprehensive Analysis of Non-ischemic Cerebral Enhancing (NICE) Lesions Following Flow Diversion for Intracranial Aneurysms: Implications and Research Imperatives

    Dear Editor,

    I am writing to express my appreciation for the recent publication of the multicenter study entitled "Non-ischemic cerebral enhancing (NICE) lesions after flow diversion for intracranial aneurysms: a multicenter study." The investigation sheds light on a relatively underexplored aspect of neuroendovascular procedures and provides valuable insights into the incidence, clinical presentation, and potential device-related factors associated with NICE lesions.

    The study, conducted across eight centers, presents a meticulous analysis of 15 patients with NICE lesions following flow diversion. Notably, the reported incidence of 1% raises concerns given its contrast with rates observed in previous studies on various neuroendovascular procedures. This finding underscores the need for a closer examination of the potential risks and implications of flow diversion in the context of NICE lesions.

    One of the notable contributions of this study is the identification of a concerning accumulation of NICE lesion cases associated with specific product lines—Pipeline devices and Derivo devices. While the study rules out the device material itself as the exclusive culprit, the hypotheses put forth regarding mechanical properties, vendor-specific catheters, and the call for further bench testing and transparent disclosure of technical details add depth to the discussion. This nuanced exploration of potential contributing factors sets the stage for more...

    Show More
  • Cost-effectiveness of Endovascular treatment for Acute Ischemic Stroke in Japan

    Dear Editor,
    I am writing to provide a thorough assessment of the recently published study titled "Cost-effectiveness of Endovascular Therapy for Acute Stroke with a Large Ischemic Region in Japan: Impact of the Alberta Stroke Program Early CT Score on Cost-effectiveness" [1]. While the study addresses important aspects of the economic implications of endovascular therapy (EVT) for acute ischemic stroke (AIS) in Japan, my analysis aims to delve deeper into specific methodological considerations and discuss the applicability of the study's findings in a broader context.
    The primary focus of the study is the cost-effectiveness of EVT based on the Alberta Stroke Program Early CT Score (ASPECTS) for patients with AIS involving a large ischemic region. The conclusion that EVT is cost-effective for individuals with ASPECTS of 3–5 as determined by treating neurologists using MRI raises questions about the reliability and consistency of ASPECTS as a determinant of cost-effectiveness. Furthermore, the study acknowledges the variability in incremental costs and quality-adjusted life years (QALYs) associated with EVT in Japan compared to other countries. The higher incremental costs in Japan, attributed to the specific stroke care system and the frequent use of combined techniques and MRI, present challenges when applying these findings to healthcare systems with different cost structures and resource allocations.
    The study acknowledges the variabilit...

    Show More
  • Delving Deeper into the Correlation between Liquid Embolic Surface Area and Chronic Subdural Hematoma Resolution in Middle Meningeal Artery Embolization

    Dear Editor,
    I would like to commend the authors for their insightful study titled "Liquid embolic surface area as a predictor of chronic subdural hematoma resolution in middle meningeal artery embolization" [1]. The investigation into the correlation between liquid embolic surface area (LEA SA) and chronic subdural hematoma (cSDH) resolution in the context of middle meningeal artery embolization (MMAE) presents valuable contributions to the field. The authors employed a meticulous approach, retrospectively collecting data from 74 patients who underwent first-line MMAE with ethylene vinyl alcohol (EVOH) and utilizing 3D segmentation to quantify LEA SA. The observed correlation between greater LEA SA and enhanced cSDH resolution rates at 3 months and 6 months post-embolization is a noteworthy contribution to the field.
    The study's strength lies in its unique focus on a patient group undergoing first-line MMAE for cSDH, a subset that has been relatively underrepresented in the existing literature. The authors rightly acknowledge the potential selection bias in this group, considering patients with greater midline shift or poor neurologic exams might be directed immediately to surgical treatment. Nonetheless, the results support the utilization of upfront MMAE in patients with riskier surgical profiles due to comorbidities or borderline radiographic and clinical features.
    However, it is essential to scrutinize the limitations outlined by the a...

    Show More
  • Analysis of Intracranial Hemorrhagic Complications in Anterior Circulation Endovascular Thrombectomy

    To the Editor,

    I am writing to discuss the recent article titled "Incidence of intracranial hemorrhagic complications after anterior circulation endovascular thrombectomy in relation to occlusion site: a nationwide observational register study" (1). This study provides valuable insights into intracranial hemorrhage (ICH) as a potential complication of endovascular thrombectomy (EVT) in patients with anterior circulation vessel occlusion stroke. The authors conducted a comprehensive analysis, considering different occlusion sites and their associations with the incidence and severity of ICH, specifically symptomatic (sICH) and non-symptomatic (non-sICH) cases. While the study presents crucial findings, it is essential to discuss its implications and limitations.

    The study's key finding of a 4.5% incidence of sICH after EVT for anterior circulation vessel occlusion stroke is consistent with previous research in this area. The recognition of differences in the frequency and severity of ICH across occlusion sites, particularly in the internal carotid artery (ICA), middle cerebral artery's first segment (M1), and the M2 and beyond, is a significant contribution to our understanding of EVT outcomes. The study suggests that ICA occlusions, despite their lower overall hemorrhage frequency, tend to result in more severe ICH, including intraventricular hemorrhages and space-occupying intracerebral hemorrhages. This observation aligns with prior stud...

    Show More
  • Middle meningeal artery embolization versus surgical evacuation, How real-world are we?

    Dear Editor,

    We have read with great interest the article entitled “Efficacy and mid-term outcome of middle meningeal artery embolization with or without burr hole evacuation for chronic subdural hematoma compared with burr hole evacuation alone” by Onyinzo, C., et al, published in Journal of NeuroInterventional Surgery (2021).

    The article has compared the Surgical and Endovascular management of (CSDH) in the elderly population who are at high risk to developing (CSDH) due to their co-morbidities and the use of anti-thrombotic agents.
    It is noted that the anti-thrombotic agents were stopped to both management arms. Upfront Middle meningeal artery embolization without stopping the anti-thrombotics, might be a strategy to mitigate the risk of cardio-embolic events.
    In regards to the evaluation of the patients outcome, clinical parameters did not include motor assessment, which is a significant factor to favor a rapid surgical evacuation in these delicate patients.
    As to the radiological follow up, cured (CSDH) were defined in the article with a thickness less than 10 mm. This is debatable as there are a lot of variations exists for this population in regards to brain volume and brain elasticity.
    Finally, we would like to point out the timing of the follow up was not defines for all patients. This need to be harmonized better to a well-defined follow up timeline. Joyce, MD, et al,2 a suggested a time frame of 6 weeks after the treatment...

    Show More
  • The number of clinical events per variable in logistic regression analysis

    Dear Editor,

    I read with interest the paper by Pierot et al [1]. They conducted a prospective study to examine factors of delayed thromboembolic events in 335 patients after coiling of unruptured intracranial aneurysms. The number of delayed TEEs was 8. The adjusted odds ratios (95% confidence intervals) of autosomal dominant polycystic kidney disease and post-procedure aneurysm remnant at procedure completion for delayed TEEs were 27.3 (3.9 to 190.2) and 9.9 (1.0 to 51.3), respectively. They understand the lack of statistical power in the multivariate analysis and did not intend to examine the causal association. I present a comment regarding the number of events in logistic regression analysis.

    The limitation in the total number of events for logistic regression analysis was simulated to improve statistical power [2]. In addition, Peduzzi et al. evaluated the effect of the number of events per variable (EPV) on the outcome in logistic regression analysis [3], concluding that the number of EPV less than 10 has some problems for the prediction of dependent variable. There is an opinion that EPV value less than 10 is also acceptable to evaluate the association by logistic regression analysis [4]. Pierot et al. observed 8 events, which was not appropriate for multivariate analysis even for examining the association instead of prediction in a prospective study. I think that wide ranges of 95% confidence intervals may reflect unstable estimates in a logistic regr...

    Show More

Pages